Ryan Liu Email & Phone Number

San Jose, Ca, Us

- Performed exploratory data analysis on varying sports datasets to uncover trends and patterns to analyze the relationship between variables and draw meaningful conclusions, and translated these results into clear visualizations using Matplotlib and Seaborn. - Developed multiple different prediction and classification machine learning models, like linear regression and random forest models, to analyze data, and so far have authored 4 published articles summarizing findings and insights… Show more - Performed exploratory data analysis on varying sports datasets to uncover trends and patterns to analyze the relationship between variables and draw meaningful conclusions, and translated these results into clear visualizations using Matplotlib and Seaborn. - Developed multiple different prediction and classification machine learning models, like linear regression and random forest models, to analyze data, and so far have authored 4 published articles summarizing findings and insights about different sports related questions and topics. Show less

- Developed and implemented sales forecasting and machine learning models for forecasting of company sales, including SARIMAX, ARIMA, Random Forest, and Linear Regression. Leveraged Snowflake for data cleaning and time-based feature engineering, and analysis of significant variables impacting sales to improve performance, including customer retention, potential sales opportunities, and common buyers. Deployed forecasting model, and is currently being used by the forecasting team to forecast… Show more - Developed and implemented sales forecasting and machine learning models for forecasting of company sales, including SARIMAX, ARIMA, Random Forest, and Linear Regression. Leveraged Snowflake for data cleaning and time-based feature engineering, and analysis of significant variables impacting sales to improve performance, including customer retention, potential sales opportunities, and common buyers. Deployed forecasting model, and is currently being used by the forecasting team to forecast sales on a weekly and monthly scale up to 8 weeks ahead, and has achieved results within 13% of actuals.- Utilized SIGMA computing for exploratory data analysis to create dynamic visualizations and detailed reports to effectively communicate insights regarding various features of the data and model performance metrics to stakeholders and upper level management. Show less

Fremont, California, United States

- Implemented and optimized various machine learning classification methods, developing random forests and neural networks to classify battery voltage cycles, and conducted feature engineering to extract relevant features from cycle data and improve classification accuracy. Project achieved 92% accuracy and is currently implemented into their product, helping the engineering team of 10 streamline testing of different materials for their batteries, lowering testing time by 5 minutes.-… Show more - Implemented and optimized various machine learning classification methods, developing random forests and neural networks to classify battery voltage cycles, and conducted feature engineering to extract relevant features from cycle data and improve classification accuracy. Project achieved 92% accuracy and is currently implemented into their product, helping the engineering team of 10 streamline testing of different materials for their batteries, lowering testing time by 5 minutes.- Provided support with database enhancement work, utilizing k-means clustering and DBSCAN to perform data segmentation and clustering analysis, and applied principal component analysis, to reduce dimensionality of high-dimensional datasets. Cleaned datasets and removed missing and incomplete data, improving data completeness by 8%. Show less

- Developed hybrid machine learning architecture consisting of convolutional neural networks and logistic regression to analyze facial features and social behavior in order to diagnose early childhood autism, achieving 87% accuracy. ​​Project accepted to present at an International Machine Learning and Artificial Intelligence Conference for Healthcare Applications. Placed 1st in Synopsys Championship, 3rd in California Science and Engineering Fair. Additionally received UCLA Brain Research… Show more - Developed hybrid machine learning architecture consisting of convolutional neural networks and logistic regression to analyze facial features and social behavior in order to diagnose early childhood autism, achieving 87% accuracy. ​​Project accepted to present at an International Machine Learning and Artificial Intelligence Conference for Healthcare Applications. Placed 1st in Synopsys Championship, 3rd in California Science and Engineering Fair. Additionally received UCLA Brain Research Institute Award and Regeneron Biomedical Science Award. - Published two papers in the Cornell arxiv, endorsed research repository: “Developing a New Autism Diagnosis System Based on Social Behavior Symptoms’ and Facial Features Extracted from Qualified Home Videos” (March 2021), “Proposing a System Level Machine Learning Hybrid Architecture and Approach for a Comprehensive Autism Spectrum Disorder Diagnosis” (September 2021) Show less

Bachelor'S Degree

Education record